Author: Zixi.eth

 

1. MEV: Maximum Extractable Value

Introduction

MEV originally refers to Miners extractable value. It is applied to POW and is a measure of the profit that miners can obtain through their ability to arbitrarily add, exclude, or reorder transactions within the blocks they produce. MEV is not limited to miners in blockchains based on proof of work (PoW), but also applies to validators in proof of stake (PoS) networks. Maximum extractable value (MEV) refers to the maximum value that a validator can extract from block production that exceeds the standard block reward and gas fee by adding, deleting, and changing the order of transactions in a block under POS. Now MEV generally refers to maximum extractable value.

Now that Ethereum has transitioned to POS, any user who stakes 32ETH can become a validator. At the time of each block, a validator will be randomly selected as a block proposer to package and execute the transaction into a new state, and then the proposer will package the block to other validators for verification. From the above description, it seems that the validator should find a way to arrange the transaction order of the block in order to make a profit. But in fact, most MEVs are arranged by Searcher. This is because Searcher needs to have special algorithm technology to search and calculate the order. Searcher submits it to the proposer through its algorithm in Mempool/private domain channel, etc., and the proposer will then package it into the blockchain. One of the manifestations of MEV is sandwich attack (clamp).

MEV has both advantages and disadvantages.

The advantages are: it can improve the efficiency of DeFi, and the searcher can quickly close the price difference. The disadvantages are that some MEVs have a great impact on the user experience. For example, the attacked users will face higher slippage and extremely poor user experience, and due to the existence of gas priority fees, the searcher will sometimes significantly increase the gas fee, resulting in network congestion and poor user experience.

Figure 1: MEV achieved by Ethereum’s top 10 DeFis over the years

The division of labor among various roles in Ethereum MEV.

Searcher is responsible for monitoring the public transaction pool and Flashbots private transaction pool, and then uses the algorithm to calculate the most profitable transaction sequence in the block, and packages it and sends it to the Builder. At this time, Searcher will give a bid to express the maximum cost it is willing to pay. This cost is only visible to the builder, not to everyone through the p2p network, thereby reducing the pressure on the chain.

Figure 2: Searcher monitors public and private trading pools to use algorithms to make quotes

Any user who downloads MEV-Boost can become a Block Builder. The Builder accepts transactions from the searcher and further selects profitable blocks from them, and then sends the blocks to the relay through MEV-Boost.

Figure 3: Builder collects quotes from different searchers

Relayer is a component that realizes PBS separation and hosts the blocks sent by Builder for Proposer. Relayer accepts the blocks transmitted by Builder and passes the most profitable block header to Validator. When Validator verifies the block header, Relayer passes the entire block header to Validator to speed up the work efficiency of Validator.

Under POS, any user can stake 32 ETH to become a Validator. Currently, Lido is the largest Validator. As long as the Validator uses MEV-Boost, it can select the most profitable block from multiple relayer proposals and charge a priority fee, and then select a Proposer from among the many Validators to produce the block.

Figure 5: Proposer (Validator) ultimately selects the most profitable block to build

MEV has several forms:

1.Dex risk-free arbitrage.

Searcher uses on-chain data for analysis and utilizes price differences + flash loans in Dex for risk-free arbitrage.

2. Settlement of loans.

The searcher queries the on-chain data at the fastest speed to determine which borrowers can be liquidated, and then submits the liquidation transaction first and charges the liquidation fee.

3. Sandwich attack

For example, if I want to spend 1 million US dollars to buy ETH on Uni, this will cause a large increase in ETH on Uni. Under the premise that the Searcher pays the highest gas priority, the Searcher only needs to monitor the transaction pool and advance the transaction order of his purchase of ETH to the transaction order of my purchase of ETH, and then after I buy it, he immediately sells it to complete the sandwich attack. Although MEV may fail, that is, the Gas priority fee is not high enough, it can be seen from the figure that the cost of failure is negligible in the total cost (failure cost + success cost).

Figure 7: In MEV, the cost of failed MEV is very small

 

2. Flashbots brings MEV track?

 

Flashbots is a research and development company that aims to mitigate the negative externalities caused by MEV (such as on-chain congestion). Flashbots has launched several products, such as Flashbots Auction (with the Flashbots Relay), the Flashbots Protect RPC, MEV-Inspect, MEV-Explore and MEV-Boost.

Here we will focus on the two products, Auction (MEV-GETH) and MEV-Boost. Before the Flashbots Auction, for example, in the DeFi Summer of 2020-early 2021, the surge in Ethereum usage brought about great negative externalities, such as high gas and Ethereum congestion. This is because in the conventional trading pool in the past, users broadcast the gas bid fee p2p to all nodes, and then miners (now validators) calculate the most profitable blocks. This open bidding method will result in high gas, and all ordinary retail investors will also bear high gas, resulting in a poor user experience.

In addition, since the auction failed (i.e. insufficient gas was paid), the transaction will also be restored on the chain, occupying a certain block space, which will eventually lead to a waste of block space and reduced income for miners (validators), a lose-lose situation. Therefore, Flashbots created Auction to alleviate the above problems.

Auction(MEV-GETH)

Auction provides a private transaction pool + privacy bidding block auction mechanism, allowing validators to outsource the most profitable block building work without trust. In this private privacy transaction pool, searchers can communicate privately and do not need to pay fees for failure. MEV-Boost is an open source middleware developed by Flashbots and runs on the chain Validator to complete the Block building business, thereby realizing the separation of Proposer and Builder under POS. PBS has not yet been implemented, so it will lead to economies of scale and centralization, that is, the larger the mine, the more capable it is to develop algorithms to improve its own Search capabilities; but PBS has now been written into the development history of Ethereum.

MEV-Boost

MEV-Boost links multiple relays to select the most profitable blocks and hand them over to the Validator. MEV-Boost currently has an adoption rate of over 90%. However, as more and more users participate in MEV, the competition for MEV has become fierce. The profit of Seacher has begun to decrease, and the profit of Producer has begun to increase. From the current cumulative data, 64% of the total profit of MEV is still occupied by the algorithm-dominated Searcher. In addition, MEV also exists on other chains, such as BSC, Avalanche, etc. The competition on other chains is not fierce, but the ceiling is relatively low.

Figure 8: Searcher revenue gross profit margin is around 64%

SUAVE

One of the current trends is builder centralization, with CR5 reaching 84.29%. Builders will gradually become centralized because they have exclusive order flows (because the Builder has preset privacy functions or the Builder is more likely to choose transactions sent by the user) and can cross chains (users can not only send me Ethereum transactions, but also send me transactions from other chains for me to package). Builder centralization will lead to the failure of Ethereum POS decentralization. Therefore, in order to solve the above problems, Flashbots developed SUAVE (the Single Unified Auction for Value Expression).

Figure 9: Builders are becoming more centralized

SUAVE is an independent blockchain, but it can be understood as a plug-in that can be inserted into any other chain, acting as the Mempool and builder of other chains. SUAVE separates the Mempool and builder of all chains from other roles to achieve professional management, thereby improving the efficiency of the entire chain (if multiple chains are not set up, the builder on a single chain will be gradually eliminated). SUAVE achieves a win-win situation, that is, the blockchain itself is more decentralized, validator income is maximized, searcher/builder can set preferences and potential income becomes larger, and users can also conduct private transactions at the cheapest price.

Figure 10: SUAVE architecture

Figure 11: SUAVE can achieve cross-chain MEV

Figure 12: 2020-2022 MEV revenue breakdown

 

3. Prospects of MEV track

 

MEV Market and Profit Margins

According to Flashbots statistics, MEV has accumulated 70K ETH, or $133 million in revenue, from 10 top DeFi projects on Ethereum in 2022. Compared with the 188K ETH and $475 million in revenue during the bull market in 2021, the reason is that the on-chain transaction turned bearish (the total chain dex transaction volume dropped from $1575B in 2021 to $1255B in 2022), the on-chain leverage ratio decreased (liquidation on-chain liquidation is also one of the sources of MEV's profit, but due to excessive decline, the on-chain leverage ratio has been reduced), etc.; but the overall gross profit margin increased from 61% in 2021 to 65% in 2022, which is due to the promotion of Flashbots, which increased the searcher gross profit margin. MEV is highly dependent on on-chain activity/transaction volume, which is greatly affected by the market. For example, in the bull market of 2021, the overall revenue ceiling was $476 million.

Figure 13: The bull market is full of opportunities on the chain, and the MEV market is large, but the gas fee is high and the profit margin is low;

The bear market has low trading volume and a small MEV market, but there is no more expensive Gas War and the profit margin is high.

MEV is a track with strong cash flow

MEV is a track with strong cash flow, and it is highly positively correlated with the cycle (on-chain transaction volume, etc.). We compare the bear market data in 2022 and select DeFi, which is also a strong cash flow business, for comparison. Without discarding the cost of paying miners, we get here that all Searchers in 2022 will earn $133.7 million on the top 10 DeFi projects on Ethereum, which is a huge income in the entire blockchain industry. Although it is not as good as the top projects such as opensea, it is much higher than the income of projects such as dydx, pancake, convex, maker, synthetix, etc. In addition, this $133 million only includes the top 10 DeFi projects on Ethereum, and does not include the income of other DeFi and Layer2+other POS Layer1 on Ethereum. Compared with several other comparable competitors selected, the weighted PS is calculated to be 12.43. Finally, the weighted PS is used to calculate the corresponding reasonable valuation of MEV Searcher, which is about $1.662 billion. Therefore, considering the entire Ethereum chain + Layer2 + other POS Layer1, the ceiling will be much greater than 1.6 billion US dollars.

Figure 14: Comparison of MEV revenue in Ethereum’s top 10 DeFis and other dapps in 2022

Figure 15: In 2022, MEV Searcher's revenue is second only to OpenSea and better than Dydx, ranking high among all blockchain projects

 

Summarize

In summary, MEV is a rare underlying track in the blockchain with strong cash flow, strong transaction correlation, high income but relatively low risk.

MEV is similar to quant and market makers in some strategies, but because it does not bear any counterparty risk, MEV is more stable than quant and market makers (at least it will not be liquidated like some market makers when the market changes drastically). MEV's strategy is more rigid, but the risk is also lower. The most typical manifestation is to do clamp arbitrage in DEX. Judging from the performance of Searcher in 2022, the MEV revenue brought by the top 10 dapps on Ethereum is 133.7 million US dollars. After deducting about 1/3 of the miners' gas fees, the revenue is about 87 million US dollars. Together with other Layer1 and Layer2, the overall MEV revenue of the blockchain is quite high.

In addition, for other L1s represented by BSC, although the overall scale of BSC's MEV will be smaller than that of Ethereum's MEV, it is limited by the lack of a unified bidding system and the lack of fierce competition, so the overall net profit margin of Searcher will be very high. In addition, we have seen MEV projects similar to Flashbots that want to unify the market on Cosmos.

From the perspective of investors, MEV is a typical equity structure project, which is similar to the investment of market makers, but there is no counterparty risk, so the overall risk is smaller. Since such companies are less likely to issue tokens, the only exit paths may be mergers and acquisitions, dividends, etc.

The core of judging the MEV project is:

1. Is the search algorithm reliable?

2. Can I quote to the node instantly?

3. Control gas costs

4. Expand other chains, etc.

Therefore, the overall technical requirements for the team are relatively high. It is not a typical BD-oriented industry and may be suitable for Chinese people.

Therefore, when investing in companies similar to MEV, if the valuations in the first and second rounds are low and the team is high, we can consider investing in them. For some of our portfolios, this may also be additional business.

For example, Blocksec has a great advantage in preemption when monitoring dark pools and blocking hackers. It only needs to improve the algorithm and take advantage of the preemption to theoretically enter the MEV track. For example, Chainbase is also doing some transaction API integration. After abstracting enough algorithms, it can theoretically write its own MEV algorithm and enter MEV based on its own nodes.